Research map / Federated learning and privacy / Personalized and heterogeneous FL
Domain Sensitive Federated Learning with Fisher-Informed Pruning
Chenchen Lin, Wenhao Yuan, Zhengji Xu, Xuehe Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add DSFP, DSR and SA modules (ablation) | FedFIP baseline · Digits | avg accuracy: 74.35 -> 76.48 | Improved |
| add DSFP, DSR and SA modules (ablation) | FedFIP baseline · Office-Caltech | avg accuracy: 55.39 -> 64.44 | Improved |
| add DSFP, DSR and SA modules (ablation) | FedFIP baseline · PACS | avg accuracy: 77.89 -> 82.02 | Improved |
| add domain-sensitive Fisher pruning alone (ablation) | no-module baseline · Office-Caltech | avg accuracy: 55.39 -> 58.78 | Improved |
| increase global sparsity ratio from 0.2 to 0.6 (ablation) | FedFIP · Digits | avg accuracy: 57.58 -> 71.99 | Improved |
| increase global sparsity ratio from 0.6 to 0.8 (ablation) | FedFIP · Digits | avg accuracy: 71.99 -> 74.44 | Improved |
| use ResNet-50 instead of ResNet-18 (ablation) | FedFIP · Office-Caltech | avg accuracy: 64.44 -> 65.86 | Improved |
| use ResNet-10 instead of ResNet-18 (ablation) | FedFIP · Digits | avg accuracy: 76.48 -> 75.52 | Hurt |
| raise local importance threshold phi (ablation) | FedFIP · Digits and Office-Caltech | accuracy | Hurt |
| use overly strong contrastive regularization (ablation) | FedFIP · Digits and Office-Caltech | accuracy | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- DapperFL — DapperFL: Domain Adaptive Federated Learning with Model Fusion Pruning for Edge Devices (NeurIPS 2024) · better on 3 of 3 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Federated Domain Generalization with Data-free On-server Matching Gradient (ICLR 2025 · similarity 0.89)
- SpaFL: Communication-Efficient Federated Learning With Sparse Models And Low Computational Overhead (NeurIPS 2024 · similarity 0.89)
- FedWon: Triumphing Multi-domain Federated Learning Without Normalization (ICLR 2024 · similarity 0.88)
- Flexible Sharpness-Aware Personalized Federated Learning (AAAI 2025 · similarity 0.87)
- Federated Unsupervised Domain Generalization Using Global and Local Alignment of Gradients (AAAI 2025 · similarity 0.87)
- Multi-Source Collaborative Gradient Discrepancy Minimization for Federated Domain Generalization (AAAI 2024 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
